The antiestrogen tamoxifen and vitamin A-related compounds, the retinoids, each have strong antiproliferative effects on breast cancer cells. In combination these agents act synergistically to further inhibit growth, representing an important means for enhancing their antiproliferative actions. This has resulted in increased efficacy from both a therapeutic and chemopreventive standpoint. The present study was undertaken to define this synergism and determine the mechanism of action of this enhanced activity in combination. The effect of 4-hydroxytamoxifen (TAM) and all-trans-retinoic acid (AT) in combination on proliferation of MCF-7 breast cancer cells was studied in vitro. It was found that TAM and AT, in combination, acted synergistically to cause a time-dependent and dose-dependent inhibition of MCF-7 cell growth. Both TAM and AT each blocked cell cycle progression throughout 7 days of treatment, but without any synergistic or additive effect on this process. TAM and AT acted synergistically, however, to stimulate apoptosis, with increased DNA fragmentation and downregulation of bcl-2 protein expression in a manner temporally equivalent to the synergistic inhibition of growth. The negative growth factor transforming growth factor beta (TGF) is secreted by these cells and was studied as a potential mediator of the synergistic effects of TAM + AT on apoptosis. While TAM but not AT stimulated TGF1 secretion, in combination TAM and AT acted synergistically to induce a 5 fold increase in TGF1 secretion over 72 hours. TGF1 alone had no apoptotic effects on these cells, but TGF1 in combination with AT acted synergistically to inhibit growth, to downregulate bcl-2 protein expression, and to stimulate apoptosis of these cells in a manner analogous to that noted for TAM + AT. Co-incubation of TAM + AT with anti-TGF antibody eliminated the enhanced apoptotic effects of the agents in combination over that noted for each agent alone. Together, these findings indicate that the synergistic antiproliferative action of TAM + AT on MCF-7 cells occurs through stimulation of apoptosis, and that this action is mediated by a novel autocrine mechanism: interaction of endogenous TGF1 with AT. Because AT is also a naturally occurring hormone, these findings may have broad implications for regulation of breast cancer cell growth. AT regulation of another important mediator of apoptosis, Fas antigen, is being studied. Fas antigen, a cell surface protein and member of the TNF receptor family, is a mediator of apoptosis in multiple cell types, but is not expressed in MCF-7 breast cancer cells. AT and interferon-gamma (IFN), in combination, act synergistically to inhibit growth of these cells, which encouraged us to examine their potential regulation of Fas-antigen expression. We found that AT and IFN in combination act synergistically to induce Fas protein expression in MCF-7 cells in a time-dependent and dose-dependent manner over 48 hours. Induction of Fas mRNA was assessed by quantitative RT-PCR. AT + IFN synergistically induced Fas mRNA within 6 hours, indicating transcriptional regulation of Fas antigen expression. Following induction of Fas antigen, treatment with Fas antibody alone resulted in modest cellular growth inhibition. Coincubation of Fas antibody with cycloheximide resulted in a significant decrease in cell viability and stimulation of apoptosis between 6 and 12 hours of treatment. This was associated with characteristic morphologic changes of apoptosis and DNA fragmentation. The requirement for cycloheximide suggests the presence of endogenous inhibitors of Fas signaling in these cells. Together, these findings indicate a novel interaction of AT and IFN, and an important mechanism for the enhancement of programmed cell death in breast cancer cells. Studies are in progress to examine the role of FLIP-S and FLIP-L, STAT1, caspase 3, and bcl-2 in the inhibition of Fas signaling and their modulation by AT and IFN, and to establish an in vivo model for Fas antigen induction and stimulation of apoptosis in breast cancer. The cellular, metabolic and molecular changes which occur in a normal breast epithelial cell in the development of a breast cancer are poorly understood and need to be defined. An understanding of these changes are important for defining the carcinogenic pathway of breast cancer, for identification of prognostic biomarkers for risk assessment, identification of biomarkers for selection and monitoring of chemoprevention agents, and for the development and evaluation of new chemoprevention drugs and chemoprevention strategies. To further define these characteristics we have developed high risk breast epithelial cell lines from normal high risk tissue from women with invasive breast cancer. These cell lines have normal breast epithelial cell morphology and express cytokeratins 14 and 18, and thus contain both basal and luminal cell types. We examined growth and metabolic regulation by retinoids, vitamin-A related compounds which are naturally occuring and which have chemopreventive and therapeutic effects for breast cancer. Studies to date have shown that these cells secrete the negative growth factor transforming growth factor beta (TGF), predominantly in the isomeric form TGF2. All-trans-retinoic acid (AT) stimulates secretion of TGF2, primarily in the latent form, and inhibits growth through blockade of cell cycle progression without stimulating apoptosis. Active TGF2 has strong antiproliferative effects on these breast epithelial cells, indicating secreted TGF may have an important paracrine or endocrine regulatory role in women at risk for breast cancer. Gene expression profiles and their regulation by AT and TGF are being be determined using a human breast gene cDNA microarray (Incyte Human UniGem 2.0 array) consisting of 8998 clusters. This will allow identification of distinctive patterns of global gene expression between these two prominent naturally occurring antiproliferative agents. Confirmation of expression differences will be confirmed with duplicate array and quantitative RT-PCR. To identify potential changes in the carcinogenic pathway of the high risk cells, gene expression profiles of high risk cells will also be compared with those of normal nonrisk breast epithelial cells using the reference cell line MCF10A as an internal standard. In this manner important differences in the progression to the high risk state can be identified, as well as characterization of their regulation by important chemopreventive modulators. Two clinical trials which compliment these studies have been developed and are in the review process. The first trial (Characterization of high risk breast duct epithelium by cytology, breast duct endoscopy, and cDNA gene expression profile, David N. Danforth, PI) provides a detailed cytologic, endoscopic, and molecular characterization of high risk breast ductal epithelium from a major source, the contralateral breast in women with an ipselateral breast cancer. In this study the ductal architecture will be defined using breast ductal endoscopy. Breast ductal epithelial cells will be collected by breast duct lavage and examined cytologically. The global gene expression profile of these breast epithelial cells will be determined by cDNA microarray using RNA amplification techniques, with validation by RT-PCR. Cell lysates of ductal epithelial cells will be prepared and the pattern of protein expression for selected proteins in the high risk epithelial cells determined using proteomics tissue lysate arrays. Gross genomic alterations present in high risk breast epithelial cells will be determined by comparative genomic hybridization. Normal ductal epithelial cells from women not at increased for breast cancer (Gail model index < 1.67%) will be studied for comparative purposes. A second clinical trial (Establishment of normal high risk breast epithelial cell cultures and a high risk cell line and tissue repository from breast tissue from women at high risk for breast cancer, David N. Danforth, PI) is designed to promote and facilitate the analysis of high risk breast tissue by making available multiple types of high risk breast epithelium. Breast tissue from each of the major categories of breast cancer risk, including normal tissue adjacent to a breast cancer, normal tissue in the same breast at a site distant from a breast cancer, from the contralateral normal breast, from women with a strong family history of breast cancer (including BRCA1 or BRCA2 mutation carriers), from women without a history of breast cancer but a Gail model 5-year risk estimate of breast cancer > 1.67%, or from women with prior definitive mediastinal irradiation for lymphoma will be collected from surgical specimens. Epithelial cell lines will be developed from each respective high risk tissue. Multiple high risk cell lines and their respective tissues will be used to establish a High Risk Cell Line and Tissue repository. This material will be used to further define the carcinogenic pathway including cellular, metabolic and molecular characteristics of high risk breast epithelial cells, for cDNA and tissue microarrays, for comparative genomic hybridization studies, and to develop a drug discovery program for chemoprevention agents using a cell line screen of high risk breast epithelial cells. This diversification of cell line and tissue material will also provide excellent complimentation to ongoing studies examining mechanisms of action of chemopreventive agents against high risk breast epithelial cells, and to the clincial study described above characterizing contralateral high risk breast duct epithelium.